On addressing the security and stability issues due to false data injection attacks in DC microgrids - An adaptive observer approach

This paper proposes an observer-based methodology to detect and mitigate false data injection attacks in collaborative DC microgrids. The ability of observers to effectively detect such attacks is complicated by the presence of unknown non-linear constant power loads. This work determines that, in t...

Descripción completa

Detalles Bibliográficos
Autores: Cecilia, Andreu, Sahoo, Subham, Dragičević, Tomislav, Costa Castelló, Ramon, Blaabjerg, Frede
Tipo de recurso: artículo
Estado:Versión aceptada para publicación
Fecha de publicación:2022
País:España
Institución:Consejo Superior de Investigaciones Científicas (CSIC)
Repositorio:DIGITAL.CSIC. Repositorio Institucional del CSIC
OAI Identifier:oai:digital.csic.es:10261/296046
Acceso en línea:http://hdl.handle.net/10261/296046
Access Level:acceso abierto
Palabra clave:Adaptive observer
Constant power loads (CPLs)
Cyberattacks
Cyber-physical systems
Dc microgrid
Distributed estimation
Nonlinear observers
Resilient controller
Descripción
Sumario:This paper proposes an observer-based methodology to detect and mitigate false data injection attacks in collaborative DC microgrids. The ability of observers to effectively detect such attacks is complicated by the presence of unknown non-linear constant power loads. This work determines that, in the presence of unknown constant power loads, the considered attack detection and mitigation problem involves non linearities, locally unobservable states, unknown parameters, uncertainty and noise. Taking into account these limitations, a distributed non linear adaptive observer is proposed to overcome these limitations and solve the concerned observation problem. The necessary conditions for the stability of the distributed scheme are found out. Moreover, numerical simulations are performed and then validated in a real experimental prototype, where communication delay, uncertainty and noise are considered.